Friday, March 07, 2003
I'm taking the weekend off and staying away from the 'Net until Monday. I'm going to the book fair at the Brabanthal tomorrow and my kitchen has become too big a mess to ignore any longer.

Plus, I've been reading Hobsbawm's Age of Extremes and David Brock's Blinded by the Right for two weeks now without finishing either one, and I'd like to move on to the rest of my reading list. So, everybody have a good weekend.

I am still in the middle of debates about evolutionary psychology. *sigh* It starts to get old. Anyway, I have promised here a sort of J'accuse for EP.

At the core of EP is a method called "adaptationism." Adaptationism looks at characteristics - physical, mental, behaviour, even social - and asks what advantage they might serve the organism or organisms that possess these characteristics. It is a method which is simple enough, requiring no experiments or equipment and only a creative mind.

Let me take a fairly simple and uncontrovertial example of a human behaviour: walking. Pretty much everyone over the age of three who does not have a clear physical disorder of some kind can walk, and the way we walk is basically the same all over the world. Furthermore, the adaptive advantage of walking is also pretty clear: we can move around and still use our hands. The adaptationist program would conclude from this that people have an instinct for walking, that it is in our genes. Here, one might well claim, is an excellent example of genes producing very complex behaviour. People must be born with an inbuilt knowledge of some sort of how to walk.

However, we should take a look at this conclusion in brighter light. The adaptationist method offers us no insight whatsoever into the genes themselves. It has no program for identifying the genes responsible for a behaviour. Actually identifying genes is not held to be a high priority and is not considered strictly necessary in order claim that walking has a genetic cause. And that's a good thing, because identifying real genes is nearly impossible without a physical and usually biochemical theory of how the identified characteristic actually comes to be realised. The adaptationist method also offers us no insight into the physical realisation of a behaviour. It does not tell us anything about the neurology of walking, or the biophysics of human bodies. It has no advice to offer to clinicians - in the case of walking, people like physical therapists. It does not tell us how this inborn knowledge is implemented in brains, or how humans do the complex calculations that enable us to keep our balance. It does not shed any light on how toddlers develop walking abilities. In short, it has nothing to offer people who study walking except the contention that walking is inherent in our DNA.

Fortunately, the people who do study walking are - I think overwhelmingly by now - agreed that walking is not instinctive behaviour at all and that infants are not born with knowledge about how to walk. The most widely held conclusion about walking is that standing upright and maintaining balance are simply the optimal solution to how to move around, if you have a body shaped like a human body and live on the surface of a planet. This position does offer insights into child development. We see children learning how to walk and experimenting and failing. We see them try different strategies to balance themselves, and we see the awkward, overcontrolled motion of toddlers. All these observations make sense in light of this non-hereditary theory of walking. Physical therapists have taken considerable insight from this theory. Now, when someone suffers from a stroke and has to relearn how to walk, we know that they relearn better when we offer them mechanical supports for their weight, just like a parent who helps hold a child up while he moves his legs. This program is also beginning to offer very valuable insights into how balance calculations are done. I have seen robots equipped with neural networks actually learn how to stand up and move around with only a single "leg."

This other approach admits that walking behaviour has a lot to do with the shape of our bodies and human biology. But, it tells us that looking for a genetic influence on this behaviour is a complete waste of time. In this context, does it make any sense at all to think about genes in trying to understand walking? The adaptive advantage of walking is only very indirectly related to genes and changing the environment - for example raising a child in space - would utterly obliterate walking behaviour.

The part of social science I know best is linguistics, and much of the EP programme got its start in the world of Chomskyan structural linguistics. Chomsky argued in the early 60's that language must be a behaviour rooted in inborn knowledge. He advanced a number of arguments in support of that end. None of those arguments have stood the test of time very well. Where nativist linguistics starts now is by making the same class of argument I offered as an adaptationist explanation for walking. They look at the near-universal use of language in human society, where the exceptions are only people who have easily identified physical or developmental disorders. They see this behaviour as very complex, and yet basically similar across very diverse populations. They see children learn language in most cases during roughly the same part of their life, and they notice that this behaviour is very robust. Furthermore, it takes very little imagination to see the reproductive fitness advantages of social structures where highly specific information can be communicated.

However, the adaptationist/nativist program has offered no insight into the neurobiology of language. It has not managed to identify genes or biochemical mechanisms that underlie language. It has not offered any useful knowledge to translators, lexicographers, language educators or people interested in natural language processing. All the progress made in those fields have come from people whose work either has no bearing on the central theses of language nativists or denies at least some nativist claims outright.

And, people outside the nativist community largely believe language is a lot like walking. Given the physiology of our senses, the structure of our mouths, throats and lungs, the properties of the human nervous system in general, and the structure of the environment in which children are immersed, language is simply the optimal solution to the problem of communication, and virtually every infant discovers it unless they are prevented from doing so by some serious physical condition. If this is the case, then there is nothing in the genome at all that specifies linguistic behaviour, or even has any sort of direct influence on it except in the trivial case where it interferes with the form of our bodies. It makes no sense to talk about innate linguistic knowledge or a language "instinct."

This programme has - over the last decade especially - shed a lot of light on language. Every meaningful advance in natural language processing since 1980 has come from some kind of interactionist or empiricist theoretical base, not from attempts to uncover innate knowledge about language. All meaningful work in lexicography, translation studies and language education has been predicated on the idea that language is firmly grounded in time and place and that it is part of the social structure of the culture in which language is spoken. No linguistic universal has ever been found other than those trivially associated with the physical restrictions of bodies and limited working memory.

This same class of criticism applies to claims of genetic influence on behaviour across the board. Even in medicine, the adaptationist programme has - contrary to what Dennett claims - been of extremely little value to clinicians. It has cast very little light on the physical mechanisms so important to therapists, and when it has in every case we have had a phyical, biochemical theory of the phenomena before a real genetic influence or adaptationist account had been proposed.

Were EP really merely the study of genetic influences on behaviour, as mediated by bodies, cultures and damn near everything in the universe, I would have no objection to it per se. Such a programme, however, would look very different than what is presently on offer. It would be a programme for study of the biological roots of behaviour, alleging that a behavioural characteristic is genetic and genetically adapted only after establishing at least a plausible biological account.

What I deny most vehemently is that adaptationist thinking and drawing strong conclusions about cause from statstical analysis of the distribution of behaviours can have any value at all in the absence of an insightful theory of how some particular behaviour comes to be. This kind of theory is called a pathological theory. If you do not have any idea how something can come to be, you can not attempt to identify causes or even influences, which are really just causes with reduced rhetorical force.

The failure to understand this point is essential to the methods of EP. To give any kind of consideration to physical bodies in thinking about behaviour is, to EP advocates, to accept that phenotypes influence behaviour, and phenotypes are influenced by genotypes. Therefore, genotypes influence behaviour, QED.

But bodies are not phenotypes. Bodies are real, physical bodies that occupy space. Phenotypes are the expression of genes. Phenotypes are defined as directly caused by genes, even when they may vary for non-genetic reasons. The form of our bodies is not simply a collection of phenotypes, it is the end result of a process of physical development in which genes play a role that varies depending on the state of the body and the world around it. There is no gene, and no set of genes short of perhaps the whole genome, that specify legs in the same way that there is a gene that specifies eye colour, and there is no set of genes we could extract from the human genome and implant in a dog that would cause the dog to walk upright. Without a theory of how the body develops, and a theory of how the form of the body leads to behaviour, to merely say that something has genetic influences, and to say that something is such-and-such percent heritable is not only useless, it can actually be harmful. Yet, this is the core of every EP claim that I can think of except those that make even more direct claims about innate behaviour and knowledge.

Stripped of adaptationist thinking, "evolutionary psychology" ceases to be evolutionary. It becomes merely biological psychology and conventional sociology. Stripped of its ability to associate behaviours to genes without having to present pathological theories, it looses everything that distinguishes it from ordinary social science. The grand conclusions advanced by its luminaries loose all force. WIthout this veneer of scientism, they represent nothing but political rhetoric and should be judged in the same way as all other political rhetoric.

Wednesday, March 05, 2003
The Unloved American: Two centuries of alienating Europe.

This New Yorker article sets European Ameriphobia into historical perspective. It's reasonably just, although I dislike rhetoric that ascribes personal emotions to whole states. What I'd like to see is a companion piece on the history of American Europhobia.

Following a discussion on Matthew Yglesias blog, I promised to put up a book list for a more mainstream view of the cognitive sciences and linguistics.

Here's what I've got so far. I'll admit it's not fully thought out yet and includes one book I haven't read and one book which is not intended to be read by people without a background in linguistics. I tried not to include difficult professional pieces and to keep it all to books in English.

I intend to look for a set of articles and shorter pieces that I can either link to or steal that offer a shorter perspective,but I haven't time this morning. I've also avoided getting into heavy math or computer science.

First, cognitive linguistics. Cognitive linguistics is one of those "imperialist paradigms" that tends to colonise different areas of discourse. It's core value is the rejection of the autonomy of syntax, envisioning linguistic knowledge as much like - and indeed dependent on - other knowledge. Approaches that could be categorised as part of this field are widespread and in all probability form the dominant framework in contemporary linguistics outside the United States and the single largest minority framework within the United States. It's sort of like the Democratic Party. Folks don't all agree on much, but they want the Republicans out.

Foundations of Cognitive Linguistics, Vol. 1 and Vol. 2. Ronald Langacker.

This is the one serious expository text I said I would include. I'm looking for more readable approaches to the basic principles of Cognitive Linguistics, but I can't really find one. It is the basic text and most cited general reference in this field.

Word Grammar. Richard Hudson. (Currently out of print)

Too bad it's out of print. It's one of the best exposition of the arguments against the core principles of generative grammar and linguistic modularity.

Women, Fire and Dangerous Things. George Lakoff and Mark Johnson.

Linguistic categorisation and its relationship to cultural anthropology.

Mental Spaces. Gilles Fauconnier.

Fauconnier makes the connection between embodied cognition and linguistic constructs.

Conceptual Spaces. Peter Gardenförs.

Gardenförs advocates an understanding of mental processes in terms of the geometry of high dimensional constructs. This is key to the connectionist perspective.

The Cultural Origins of Human Cognition. Michael Tomasello.

This is the one book on this list I haven't read, but in his other books he draws the link between embedded and situatated cognition and the Vygotskyan and Piagetan traditions in development theory. It seems likely that this will provide a more unified and coherent exposisiton of his perspective.

Beyond Modularity. Annette Karmiloff-Smith, Patricia Kitcher.

This book draws on a sort of neo-Piagetan perspective to make the case that modularity of behaviour is the result - and not the precondition - of child development.

Rethinking Innateness. Jeff Elman, Elizabeth Bates, et al.

Elman, et al make the case that neural networks can serve to explain a variety of phenomena that are often held to require innate knowledge. It has proven devastating to the "poverty of stimulus" arguments that were so initially convincing in the 60's. Comes with software and is often used in undergraduate linguistic courses.

Cognition in the Wild. Edwin Hutchins.

Hutchins follows the navigation crew of a US Navy ship, showing how cognition takes place as the product of cultural frameworks. He makes his case that cognitive activity is not the product of brains alone, but of cultural institutions like the structure of a military unit.

Semantics: Primes and Universals. Anna Wierzbicka.

Summarises Wierzbicka's work in linguistics and cultural anthropology. Wierzbicka's work is focused on lexical semantics, but extends into syntax and the cultural significance of language. Essential reading for lexicographers.

Being There: Putting Brain, Body, and World Together Again. Andy Clark.

Clark does the philosophy of situated and emboddied cognition, and the new thinking about cognition in general.

I don't agree 100% with any of these authors. Each one pretty much holds some position that I don't. Someday, I'll write my book and outline a more unified theory, but unfortunately the two most important core concepts of what I want to do are not present in this list: the debate over symbolic versus subsymbolic representations and soft computing.

Still, it's a place to start.

Monday, March 03, 2003
Enetation has gone on the fritz as of... about 5 minutes after 9pm CET, and Blogger has lost the archives again. *sigh* I guess I should cough up some money for this business.
Update: 30 seconds later the archives are back.
Update: It's 8am CET and everything seems to be okay now.

The Seven Warning Signs of Bogus Science

This is up on several blogs but I have a few thoughts to add. I've been reading Bob Park for years - he has a newsletter that goes out by e-mail every Friday, you can subscribe by clicking the "What's New" link on the sidebar - and in this article he lays out in executive-summary bullet-point form some signs that you are dealing with bad science.

  1. The discoverer pitches the claim directly to the media.

  2. The discoverer says that a powerful establishment is trying to suppress his or her work.

  3. The scientific effect involved is always at the very limit of detection.

  4. Evidence for a discovery is anecdotal.

  5. The discoverer says a belief is credible because it has endured for centuries.

  6. The discoverer has worked in isolation.

  7. The discoverer must propose new laws of nature to explain an observation.

On the whole, I agree, especially with regard to the narrower context of scientific testimony in courts, which is the main point of Park's article. Science in the courts should be a conservative science, sanctioned by as large a consensus as possible. However - and this is the important bit - progress, especially big paradigm-shift type progress, takes place at the radical edges of science. I can think of fairly important work that breaks at least some of these constraints.

  1. In cognitive science, a great deal of work is published in semi-mainstream books and often these public works are the best descritptions of a school of thought. Sometimes, I think this is a good thing. Writing to a large market makes for clearer presentation than assuming the things you think your professional audience should know. But, this is also a really bad thing, because the best writer may not be the best researcher. (If you're out there Pinker, I mean you!)

  2. There have been powerful conspiracies directed against scientific claims. Not just radical instances like Galileo and Darwin, but much more recent and less vocal cases like Minsky and Papert's 1968 book Perceptrons, which succeded in killing an important area of computer science research for over a decade despite considerable promise and interesting results.

  3. The precession of Mercury's orbit is something that we take nowadays as the first proof of Einstein's theory of general relativity, but we should note that it - like most phenomena that could confirm or disconfirm relativity - was at the very brink of detection during the period when Einstein devised his theory.

  4. I'm with Park here, except I'm not clear what he means by anecdote. It's true that whole scientific frameworks are not deduced from single experiments.

  5. Again, I'm mostly with Park here. I would put it differently though: an argument from a default hypothesis isn't worth a hill of beans. However, not every human practice is the result of careful experiment or analysis, many are simply a matter of cultural inheritance that we accept without questioning. It probably can't be any other way. We shouldn't make the claim that those practices are scientific just because we do them, but claiming that some practice isn't scientific isn't a good reason to stop either.

  6. Quite a few theoretical breakthroughs have occurred in isolation. Einstein was quite famous for his Gedankenexperiments, conducted in a purely mental fashion. Where Park is - I think - going with this is that there is a cultural standard for scientific authority - a sense that a scientist is someone other scientists recognise as such. Those too far removed from that social acceptance don't get the benefit of the doubt. I agree with this wholly, but I'm not sure Park would agree with how I phrased it.

  7. I'm unclear where Park is going with this one. Proposing new laws of nature is sometimes the whole point of a line of research. Now, Park does say that the new laws have to be consistent with the old observations. True enough, but not very helpful.

Park is concerned about a number of areas of, well, he would call it "bad science." Some of them concern me too. He has railed against the ISS and the Space Shuttle as a waste of money, and against SDI (= missle defense) as based on hopelessly bad scientific and engineering practices. These matters concern me too, because these are essentially political battles with important social outcomes.

Where I find him less interesting is on "pseudo-science." It's not that he isn't right about perpetual motion machines and homeopathy, it's that he wants to outline clear principles to distinguish science from non-science and teach those principles to the public. I think that is about as useful as teaching them to distinguish good theology from bad. Science is essentially a culture, and being a scientist is in part acceptance of that culture and its values. Those values aren't simple or clear, and they are not rational principles that form a totally consistent system of thought.

We can offer the public the chance to accept scientific culture's values - something that is not happening now - and many people will inevitably refuse to accept them. Furthermore, some of those who do accept them will reject some part of what most people take those values to mean and sometimes they will be right to do so.

What I fear would come from trying to teach scientific principles as such is turning science into a new sort of theology with scientists sqabbling about it like so many priests arguing over the number of angels who can sit on the head of a needle. There are already too many signs of this, especially in my field - linguistics - where every major conflict of methodolgy or philosophy turns into an argument over the nature of the scientific method.

And perhaps worse, I fear turning science into a new church for the masses who don't imagine themselves to be scientists. Park knows that it is the claim to scientific authority that makes modern pseudo-science so powerful. The right way to answer this challenge is to teach the public to look at scientific claims critically, even though that means a reduction in the authority of scientists and an acceptance of scientific disputes as poltical to an uncomfortable degree.

The Two Faces of Fred: A Journalistic Moral Tale

I have a few shameful indulgences. Well, actually I have quite a few shameful indulgences. I mean, things that I hate to admit enjoying because I know how they look. Like my nostalgic fondness for the music of Stacey Q or how I would hate to be caught dead reading Das Bild, Germany's biggest soft porn tabloid.

One of these indulgences is The Exile, a bimonthly English language "alternative" newspaper from Moscow. It's certainly an alternative you won't find anywhere else in the world. There is hardly any negative adjective that you couldn't attach to it or some part of it - misogynistic, drug-adled, libelous, irresponsible, possibly fascist - nor is there any depth of journalistic malpractice that they won't plunge to. Indeed, they seem to take a great deal of pride in knowing that if they operated in any saner country than the Russian Federation, they would either be tied up in several hundred libel trials, in prison, or simply executed by a firing squad.

For example, they are currently running a contest for the first person to bring in physical proof that TaTu, that Russian jailbait fausse-lesbian duo that's been doing so well on the pop charts are, in fact, straight. The Exile is quite famous for their regular "death porn" column, in which they pick the most disgusting and disturbing violent crime stories in Russia and summarise them in the most revolting terms they can find. They are also famous for ranking nightclubs by the price of the booze, the odds of getting laid, and your chances of getting out alive.

Part of what I like about them is that they are an example of what a truly free press would have on offer, for better and for worse. They represent every tabloid value taken to its ultimate extreme. And part of what I like about them is a sort of ruthless honesty, shorn of any pretense or politeness.

But, I also like them because at least some of what is on offer at the Exile is real journalists, who are involved in the more "legitimate" press and who have an actual sense of journalistic values, but who get to forget those values and say what they damn well please because at the Exile, no one gives a shit.

This is the case with the pseudonymous "Philby Burgess." Philby and Burgess are the names of famous Soviet spies who were part of the "Cambridge Ring" from the early 50's. His (or her) byline informs us that Philby Burgess is the pseudonym for a Western correspondent living in the former Soviet Union.

The Philby Burgess piece in the current Exile - The Two Faces of Fred: A Journalistic Moral Tale - manages, in 322 words, to malign the Washington Post and its opinion page editor, the Bush administration's rhetoric over terrorism and Iraq, the Russian government's rhetoric over Chechnya and finally the whole American press, which is held to be hypocritical in its treatment of Chechnya and the middle east.

Chechnya is exactly the sort of cause that can drive someone who is trying not to be a hypocrite nuts. One would like to support the underdog, fighting for national liberation from distant overlords. And yet, as is often the case, one finds that the good people fighting for their freedom are not really so good, nor is their national liberation necessarily very liberating. I am hard pressed to see how letting Russia fall apart will make the world a better place, and I suspect that a sovereign Chennya is just another Somalia or Afghanistan waiting to happen. But, I can't help but agree with Russians who have no great desire to go die, or see others go off to die, in a cruel little regional war at the far fringes of their country. I am as horrified as everyone else by Russian actions in the region, and I am just as horrified by Chechen terrorists who seem to be taking their inspiration from Al-Qaeda.

So, I'm asking for some help. Does anybody out there have a consistent position on both Chechnya and the middle east? How do you come to a conclusion on one that doesn't make a mess of the other, short of "kill'em all and let God sort'em out?" Bonus points will go those who are also able to maintain a rational, non-hypocritical position on Israel in the same framework.

Sunday, March 02, 2003
The Internet and American media

I found a couple of interesting things up on the web this afternoon. The first is an article about UK online news sources titled US public turns to Europe for news . I got it from a blog that I don't read very often, but I can't remember which one right now.

The threat of war in Iraq is driving increasing numbers of Americans to British and international news web sites in search of the broader picture.

According to the internet audience management and analysis company, Nielsen NetRatings, traffic to the UK's biggest news sites, BBC News Online and Guardian Unlimited, has increased dramatically over the past year. Many of these new users are from the US.

I think this is a very positive trend. Even moving from CBC and the Globe and Mail to the American media is a big shock.

The second item I'd like to point to is this White House press briefing from last Tuesday. I got this too from a blog, and have now forgotten which.

Q -- the French press is quoting actually two different diplomats from the United States State Department that -- they're highlighting that the United States is giving some sort of agreements or benefits to Colombia -- and other non-members of the Security Council --

MR. FLEISCHER: I haven't seen the story. And you already have the answer, about what this will be decided on. But think about the implications of what you're saying. You're saying that the leaders of other nations are buyable. And that is not an acceptable proposition. (Laughter.)

You can hear this on the audio track as well by advancing to minute 36. Here is Ari Fleischer, disclaiming the idea that the US would try to buy some government off, and the White House press corps laughs him out of the room. This is priceless theatre, and as soon as I find my patch cables, I'll put the relevant section up on the web in WAV format.

Both cases serve as examples of how blogging and the diverse media sources on the web have improved my ability to inform myself. I certainly would probably not have encountered either story on my own. The web may be good for something after all.